tion, sodium diethyldithiocarbamate (NaDDTC) and me- thylisobutylketone (MIBK) were used. Sensitivity: lead. 2 µg/l, cadmium 0.1 µg/l. Linear area: lead up to ...
European Society for Human Reproduction and Embryology
Human Reproduction Update 1998, Vol. 4, No. 3 pp. 301–309
Impact of heavy metals on hormonal and immunological factors in women with repeated miscarriages I.Gerhard1,3, S.Waibel1, V.Daniel2 and B.Runnebaum1 1Department of
Endocrinology and Reproduction, University Hospital of Obstetrics and Gynaecology, Voßstr. 9, 69115 Heidelberg and 2Institute of Immunology, University of Heidelberg, Germany
TABLE OF CONTENTS Introduction Observational patient group General investigations Investigation of heavy metal body load Statistical analysis Results of general investigations Results of heavy metal and laboratory findings Discussion References
301 301 302 303 303 303 304 307 308
In 111 women with repeated miscarriages, the urinary excretion of heavy metals was determined in a challenge test with the chelating agent 2,3-dimercaptopropane-1-sulphonic acid in addition to hormonal, chromosomal, immunological and uterine investigations. The heavy metal excretion was correlated to different immunological (natural killer cells, T cell subpopulations) and hormonal (progesterone, oestradiol, prolactin, thyroid stimulating hormone) parameters. We conclude that heavy metals seem to have a negative impact on ovarian as well as on pituitary function. The heavy metal-induced immunological changes may interfere with the physiological adaptation of the immune system to the state of pregnancy with the result of a miscarriage. The observed heavy metal-induced hormonal and immunological changes may be important factors in the pathogenesis of repeated miscarriages. Key words: DMPS testing/heavy metals/immunology/ infertility/repeated miscarriages Introduction Hormonal and immunological disorders have been identified as the main causes of repeated miscarriages (Gerhard et al., 3To
1981, 1996; Unander et al., 1985, 1987; Gerhard and Runnebaum 1986; Makino et al., 1992; Clifford et al., 1994). The diagnosis of immunologically caused recurrent miscarriages remains difficult and is made on the basis of excluding other common reasons. In recent years, we have been investigating the impact of harmful substances on the endocrine system and fertility in humans. It was observed in animal studies and accidental poisoning cases in humans that the increased uptake of lead, cadmium or mercury interfered with the normal pregnancy course and resulted in miscarriages, fetal malformation and stillbirth (Amin-Zaki et al., 1974; Flora and Tandon 1987; McMichael et al., 1986; Rowland et al., 1994; Colborn et al., 1996). In a previous study of infertile women, we found a significantly increased prevalence of hormonal disorders among subjects with high mercury body load. Furthermore, women with a history of miscarriages presented with a significantly greater cadmium body load than the other subjects (Gerhard, 1995a). This study was designed to investigate possible connections between the heavy metal body load and hormonal or immunological changes in women with repeated miscarriages. Observational patient group The heavy metal excretion was investigated in 111 women with a history of at least two miscarriages, median age 31 years (range 21–39 years), in addition to the usual diagnostic procedures in our department between 1989 and 1993 (Table I). Forty-three women had a history of two, 35 subjects of three and 33 women of four or more miscarriages. Eighty women complained of primary miscarriages (patient has never been pregnant before), 31 women of secondary miscarriages (delivery of a least one baby with the same partner). Seventy-nine women complained of early miscarriages (≤12th week of gestation), six women of late miscarriages (13th–28th weeks of gestation). Twenty-six
whom correspondence should be addressed Tel: +49 6221 568321; Fax: +49 6221 565433
302
I.Gerhard et al.
women had a history of both types. The mean body weight was 62.6 kg (range 45–98 kg), the median body mass index (BMI, kg/m2) 22.4 (range 17.4–37.4). Fifteen per cent of the women were housewives, 72% office-workers and 13% industrial workers. Thirty-three women (29.8%) were smokers. Twenty women smoked 20 cigarettes per day. The menstrual cycles were found to be eumenorrhoeic (interval 25–30 days) in 70%, polymenorrhoeic (interval 6 months) in 3%. All subjects were medically fit and on no medication except for nine women on thyroxine for hypothyroidism. General investigations To identify the cause of repeated miscarriages hormonal, uterine, chromosomal and immunological investigations were performed (Table I). In women with luteal insufficiency (two different determinations of luteal progesterone 450
mE/l) and/or hyperandrogenaemia [early follicular testosterone >500 pg/ml and/or dehydroepiandrosterone sulphate (DHEAS) >4500 ng/ml], the hormonal disorders were considered to be the reason for repeated miscarriages. As long-term treatment is required for thyroid disorders, these were not considered as hormonal disorders as cause of repeated miscarriages in the context of this study. The antiphospholipid syndrome (APS) was suspected in patients with a partial thromboplastin time (PTT) >30 s and/or anticardiolipin-antibodies >6 U/ml, though none of the patients fulfilled the clinical criteria, e.g. previous thrombosis, as defined by Harris (1990). Patients with various positive antibodies with uncertain pathogenetic relevance for repeated miscarriages included women who were positive for antinuclear antibodies (ANA), anti-DNA antibodies, human leukocyte antigen (HLA) antibodies or for blocking factors in cross-match. The details of the hormonal and immunological analyses have been described recently, as were the methods, sensitivity, the coefficients of intra- and interassay variance (Gerhard et al., 1991, 1993, 1997a; Daniel et al., 1996).
Table I. Investigations of the causes of repeated miscarriages and abnormal results in the total group of women with repeated miscarriages (n = 111) Investigation
n
Hormonal tests: Follicular phase: gonadotrophins (FSH, LH), prolactin, testosterone, oestradiol, DHEAS, T3, T4, TSH (basal, 30 min after TRH) Luteal phase: progesterone Uterine and tubal investigation: Hysterosalpingography, chromolaparoscopy
111
Chromosomal analysis: Karyogram Immunology: Anticardiolipin antibodies Anti-DNA Antinuclear antibodies (ANA) HLA antibodies HLA typing of the couple Lymphocyte subpopulations Cross-match Interleukin-2 receptor Immunoelectrophoresis Various: Partial thromboplastin time Full blood count and differential, biochemical profile
Abnormal results
n
Hyperprolactinaemia Hyperandrogenaemia Mixed hormonal disorder
10 8 1
Luteal insufficiency
16
Uterus septus, bicornis Uterine fibroids Partial or complete blockage of one tube
5 1 11
73
79
93 85 85 111 101 102 85 22 86 90 111
Mosaic 46XX (92%), 47XXX (8%)
1
Positive >6 GPL/ml Positive Positive Positive Matching antigens: 1 – n = 39, 2 – n = 26, 3 – n = 12, 4 – n = 6
6 2 2 1
Positive blocking factors Increased (>5000 pg/ml) Increased IgM (>2.8 g/l) Prolonged (>30 s)
4 7 11 6
FSH = follicle stimulating hormone; LH = luteinizing hormone; DHEAS = dehydroepiandrosterone sulphate; TSH = thyroid stimulating hormone; TRH = thyrotrophin releasing hormone; HLA = human leukocyte antigen.
Heavy metals and miscarriage
Investigation of heavy metal body load To estimate the heavy metal body load, a challenge test with the chelating agent 2,3-dimercaptopropane-1-sulphonic acid (DMPS) was performed as follows: After a 12 h fast, a 10 ml urine sample was collected at 08.00 h. DMPS capsules (Dimaval Heyl Co., Berlin) were given orally (10 mg/kg). A further 10 ml of urine was collected after 2 and 3 h when the maximum excretion takes place (Gerhard et al., 1992b). The concentrations of mercury (Hg), lead (Pb), cadmium (Cd) and arsenic (As) were determined in each sample. Zinc (Zn) and selenium (Sn) concentrations were measured only in the basal sample. The concentrations were set in relation to the corresponding creatinine content of the sample. The following analytical methods were used. Lead and cadmium
Atomic absorption spectrometry (AAS) with graphite tube (Perkin-Elmer 1100 B; Überlingen, Germany). For extraction, sodium diethyldithiocarbamate (NaDDTC) and methylisobutylketone (MIBK) were used. Sensitivity: lead 2 µg/l, cadmium 0.1 µg/l. Linear area: lead up to 100 µg/l, cadmium up to 6 µg/l. Coefficient of variation: lead 50 µg/l up to 5%, cadmium 13 µg/l up to 3%. Mercury
AAS cold steam technique (Perkin-Elmer 2100) in combination with flow injection system. 100 µl nitric acid was added to the urine. Reduction was performed with sodium boron hydride solution 0.02%. Sensitivity: 1 µg/l. Linear area: up to 200 µg/l. Coefficient of variation: 7 µg/l up to 8%. Arsenic
AAS hydride technique (Perkin-Elmer 3030) in combination with MHS-20-hydride unit. Direct analysis was performed in comparison to aqueous arsenic (III) standard solution. Sensitivity: 1 µg/l. Linear area: up to 40 µg/l. Coefficient of variation: 10 µg/l up to 7%. Zinc
Flame technique AAS (Perkin-Elmer 2100). Direct analysis was performed in comparison to aqueous standard solution. Sensitivity: 20 mg/l. Linear area: up to 1000 mg/l. Coefficient of variation: 650 mg/l up to 6.5%. Selenium
AAS hydride technique (Perkin-Elmer 3030) in combination with MHS-20-hydride unit. Breaking down of the samples was performed with nitric acid and hydrochloric
303
acid. Reduction was performed with sodium boron hydride solution 15%. Sensitivity: 2 µg/l. Linear area: up to 100 µg/l. Coefficient of variation: 70 µg/l up to 9%. To estimate a simultaneous body load of lead, cadmium and mercury, we used the metal score. In the total group, the values of maximum stimulated urinary excretion were divided into percentiles. Each subject scored 1 to 4 points for the urinary excretion of lead, cadmium and mercury (1 point: excretion 440/nl). Women with primary versus secondary miscarriages had a significantly higher platelet (303 versus 250/nl, P = 0.0004), total T cell (1.454, 1107–2050 versus 1.238, 788–1665/µl, P = 0.0406) and T helper cell count (1002, 670–1320 versus 753, 498–1133/µl, P = 0.0434). Subjects with late (2.09, 1.46–2.37 versus 1.59, 1.18–2.23 g/l, P = 0.06) or secondary (2.09, 1.55–2.61 versus 1.57, 1.27–2.23 g/l, P = 0.0414) miscarriages had significantly higher IgM concentrations than women with
304
I.Gerhard et al.
early or primary miscarriages. Serum selenium concentrations (normal range 73–165 µg/l) were reduced in 38.5% and serum zinc (normal range 70–150 µg/l) concentrations in 18.5%. Results of heavy metal and laboratory findings The urinary heavy metal excretion is shown in Table II. Basal cadmium excretion was significantly greater in women with primary versus secondary miscarriages (0.33 versus 0.2 µg/g creatinine, P = 0.0136). No differences in heavy metal excretion were noted in women with early versus late miscarriages, smoker versus non-smoker or in association with general parameters, e.g. BMI. A significant direct correlation between lead and age was found (Spearman correlation coefficient r = 0.22973, P = 0.02). The urinary mercury excretion was significantly associated with the number of amalgam tooth fillings (Figure 1). The number of fillings correlated more strongly with the DMPS-stimulated excretion (r = 0.68, P = 0.0001) than with the basal mercury excretion (r = 0.35, P = 0.001). Correlation between heavy metals and laboratory findings Mercury (Table III)
In women with secondary miscarriages, basal mercury excretion correlated inversely with the percentage of T suppressor cells. The lymphocyte count was directly associated with the stimulated mercury excretion. In women
Figure 1. Mean urinary mercury (Hg) excretion and the number of amalgam tooth fillings. Basal and 2,3-dimercaptopropane-1-sulphonic acid-stimulated (120 and 180 min) urinary excretion are shown.
with primary miscarriages, decreasing progesterone concentrations were noted the greater the stimulated urinary mercury excretion. Decreasing stimulated thyroid stimulating hormone (TSH) concentrations were found with increasing basal or stimulated mercury excretion in the total group and in women with secondary miscarriages. Oestradiol concentrations were directly correlated to basal mercury excretion in women with early miscarriages.
Table II. Urinary heavy metal excretion (µg/g creatinine) in the group of women with repeated miscarriages. Excretion is shown at baseline and 120 and 180 min after oral 2,3-dimercaptopropane-1-sulphonic acid stimulation (10 mg/kg body weight) Excreted substance (µg/g creatinine)
Referencea
Arsenic basal
